Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Circulation. Author manuscript; available in PMC 2018 January 3.
Published in final edited form as:
PMCID: PMC5195872

An IGF1R-dependent pathway drives epicardial adipose tissue formation after myocardial injury

Lior Zangi, PhD,1,2,3,4, Marcela S. Oliveira, PhD,4,* Lillian Y. Ye,4,* Qing Ma, MD,4 Nishat Sultana, PhD,1,2,3 Yoav Hadas, PhD,1,2,3 Elena Chepurko, PhD,1,2,3 Daniela Später, PhD,5 Bin Zhou, PhD,6 Wei Leong Chew, PhD,7 Wataru Ebina, PhD,8 Maryline Abrial, PhD,9 Qing-Dong Wang, MD, PhD,5 William T. Pu, MD,4,8, and Kenneth R. Chien, MD, PhD10,



Epicardial adipose tissue (EAT) volume and coronary artery disease are strongly associated, even after accounting for overall body mass. Despite its pathophysiological significance, the origin and paracrine signaling pathways that regulate EAT’s formation and expansion are unclear.


We used a novel modified mRNA (modRNA)-based screening approach to probe the effect of individual paracrine factors on epicardial progenitors in the adult heart.


Using two independent lineage tracing strategies in murine models, we show that cells originating from the Wt1+ mesothelial lineage, which includes epicardial cells, differentiate into EAT following myocardial infarction (MI). This differentiation process required Wt1 expression in this lineage and was stimulated by insulin-like growth factor 1 receptor (IGF1R) activation. IGF1R inhibition within this lineage significantly reduced its adipogenic differentiation, in the context of exogenous, IGF1 modRNA stimulation. Moreover, IGF1R inhibition significantly reduced Wt1-lineage cell differentiation into adipocytes after MI.


Our results establish IGF1R signaling as a key pathway that governs EAT formation in the context of myocardial injury by redirecting the fate of Wt1+ lineage cells. Our study also demonstrates the power of modRNA-based paracrine factor library screening to dissect signaling pathways that govern progenitor cell activity in homeostasis and disease.

Keywords: Epicardial adipose tissue, epicardium, Wt1, insulin related growth factor 1, myocardial infarction
Subject Codes: [130] Animal models of human disease, [147] Growth factors/cytokines, [151] Ischemic biology - basic studies


Epicardial adipose tissue (EAT), located between the epicardium and underlying myocardium, constitutes 20% of the mass of the human heart.1 EAT volume and coronary artery disease are strongly associated, even after accounting for overall body mass.13 The association of EAT volume to heart disease has been attributed to paracrine signaling between EAT and adjacent coronary vessels and myocardium.2 Despite its pathophysiological significance, the origin and paracrine signaling pathways that regulate EAT’s formation and expansion are unclear.

The epicardium is a specialized form of mesothelium, the polarized epithelium that lines the surface of many organs including the heart, lung, liver, and gut. In the developing heart, epicardial progenitors, marked by expression of the transcription factor WT1, undergo epithelial-to-mesenchymal transition (EMT) to form epicardium-derived cells (EPDCs), which migrate into the heart to form fibroblasts, smooth muscle cells, and, possibly, endothelial cells and cardiomyocytes.46 In the adult heart, the epicardium is normally quiescent, but myocardial infarction (MI) reactivates a subset of its fetal program, inducing epicardial thickening, fetal gene reactivation, and differentiation of epicardial progenitors into fibroblasts and myofibroblasts.7 The behavior and fate of injury-activated epicardium is guided by paracrine signals, as we showed in proof-of-concept experiments in which VEGF-A, delivered to myocardium at the time of MI using modRNA, redirected epicardial progenitor fate by stimulating EPDC expansion, mobilizing their migration into subjacent myocardium, and directing their differentiation into endothelial cells.8

Here, we used a novel in vivo, modified mRNA (modRNA)-based paracrine library screening approach to show that IGF1R activation is sufficient to promote EAT formation in the context of MI. In this setting IGF1R activation enhances the ability of Wt1-lineage cells to differentiate into adipocytes, and IGF1R inhibition or Wt1 inactivation blocked their MI-induced contribution to EAT. Our results establish IGF1R signaling as a key pathway that governs EAT formation and the fate decisions of Wt1-lineage cells in the context of myocardial injury.


Please refer to Expanded Methods in the Online Supplement for details.


All animal procedures were performed under protocols approved by the Boston Children’s Hospital Institutional Care and Use Committee. Wt1CreERT2, Wt1GFPCre/+, Wt1flox/flox, αMHCMerCreMer, R26Tomato, and R26mTmG alleles have been described previously.6,912 Tamoxifen (Tam, 0.12 mg/g body weight) was administered to adult mice twice weekly for 3 weeks to induce CreERT2-mediated recombination. One week after completion of Tam dosing (to allow Tam clearance), the left anterior descending coronary artery was ligated to induce MI.13 Paracrine factor modRNAs (100 μg/heart, Luciferase (Luc) modRNA serve as control) were injected into the infarct zone myocardium immediately after LAD ligation.8 ModRNA gels were applied to the heart either at the time of LAD ligation, or two weeks prior via a lower thoracotomy, as described in the text. Procedures and measurements were performed blinded to genotype and treatment group. All animals that started an experimental protocol and that survived to the measurement point were included.

Synthesis of modRNA

ModRNAs were in vitro transcribed from plasmid templates (sequences provided in the Online Supplement, Table S1) using a custom ribonucleoside blend of 3′-O-Me-m7G(5′)ppp(5′)G cap analog (6 mM, New England Biolabs), guanosine triphosphate (1.5 mM, USB), adenosine triphosphate (7.5 mM, USB), and 5-methylcytidine triphosphate and pseudouridine triphosphate (7.5 mM, TriLink Biotechnologies)as described previously 8,14,15 RNA was treated with Antarctic Phosphatase (New England Biolabs), quantitated by Nanodrop (Thermo Scientific), precipitated with ethanol and Ammonium Acetate, and resuspended in 10 mM TrisHCl, 1 mM EDTA. ModRNA was transfected into cultured cells using RNAiMAX (Life Technologies). The transfection mixture was added to cells cultured in DMEM with 2% FBS and 200 ng/ml B18R (eBioscience, San Diego, CA). ModRNA gel was made by mixing Cre modRNA (10 μl modRNA at 20 μg/μl), Lipofectamine 2000 (30 μl, Life Technologies), and 0.05% polyacrylic acid (10 μl; Sigma). The mixture was incubated for 15 minutes at room temperature to generate the gel, which was painted on the heart surface.

Isolation and culture of murine EPDCs and mesenchymal stem cells

Mesenchymal stem cells (MSCs) were isolated from adult (6–8 weeks) CFW femurs as described previously.16 Isolated cells were cultured in StemXvivo Osteogenic/Adipogenic Base media (R&D Systems) with Penicillin-Streptomycin (1:100). WT1+ EPDCs were isolated from heart explants of WT1GFPCre/+ mice 2 days after MI. Cardiac cells (non-myocytes) were allowed to expand from heart explant cultures. After 2 weeks, GFP+ cells were isolated by FACS (FACS Aria III) and plated plated in fibronectin-coated (5 ng/ml for 2 hours at 37°C) wells of a 12 well plate (70,000 cells per well).

For enhancement of adipocyte differentiation in MSCs or EPDCs, culture medium (StemXvivo Osteogenic/Adipogenic Base media; R&D Systems) was supplemented with Adipogenic Supplement (R&D Systems, 1:20). ModRNAs were transfected every 3–4 days during adipogenic differentiation. For detection of oil droplets, cultures were stained with saturated Oil red O solution (Sigma).17 To quantitate oil red O staining, plates were dried and extracted with 1 ml 100% isopropanol. After 10 minutes incubation with gentle shaking, the OD500 was recorded.

Human EPDCs

Human EPDCs were obtained from atria of adult patients undergoing heart operations under a protocol approved by The Medical Ethics Committee of the Leiden University Medical Center. EPDCs were isolated as described and passaged at most 9 times. 18


Immunostaining was performed on cryosections of hearts fixed by perfusion with 4% PFA using the antibodies listed in the Online Supplement, Table S2. Quantification of immunostaining in cardiac sections was performed using ImageJ Software.

RNA isolation and gene expression profiling

The peri-infarct zone near the apex was snap-frozen. Total RNA was isolated using the RNeasy mini kit (Qiagen) and reverse transcribed using Superscript III reverse transcriptase (Life Technologies). Real-time qPCR analyses were performed on a Mastercycler Realplex 4 Sequence Detector (Eppendoff) using SYBR Green (Quantitect SYBR Green PCR Kit, Qiagen). Fold-changes in gene expression were determined by the [partial differential][partial differential]CT method and were presented relative to Gapdh internal control. PCR primer sequences are shown in the Online Supplement, Table S3.

For microarray gene expression profiling, WT1GFPCre/+ mice underwent LAD ligation, and 7 days later the hearts were dissociated. GFP+ EPDCs were isolated by GFP FACS. RNA was isolated and used to probe Affymetrix Gene 1.0 ST arrays (n=3). Gene expression values were determined using Affy Power Tools, and the distribution of mean values were used to define the detectable gene expression threshold. Genes with gene ontology terms “receptor activity” or “plasma membrane” were manually curated to define a set of cell surface receptors.

Western blot

Western blotting was performed to measure phosphorylated IGF1R in EPDCs stimulated with IGF1 protein (PeproTech). Samples containing equal amounts of protein were separated by SDS-PAGE, transferred to nitrocellulose, probed with IGFR (pY-1161) antibody (Abcam, 1:1000) followed by donkey anti-Rabbit HRP antibody, and visualized by chemo-luminescent detection.

Statistical analyses

Values are reported as mean ± standard error of the mean. Comparisons between groups were made using Welch’s 2-tailed t-test (continuous variables) or Fisher’s exact test (proportions).


Recent lineage tracing studies indicate that EPDCs are one of the cell types that undergo adipogenic differentiation into epicardial fat1921, and that MI stimulates EPDC adipogenic differentiation20. To gain further insights into this process, we established a lineage tracing system to measure MI-induced adipogenic differentiation of EPDCs. We used a pulse-labeled genetic lineage tracing strategy, in which Wt1 regulatory elements drive epicardial expression of CreERT26. In the presence of the inducing agent tamoxifen, CreERT2 indelibly activates the Rosa26Tomato reporter, so these cells and their descendants express RFP8,19,20. We refer to these cells durably labeled by Wt1CreERT2 as Wt1-lineage cells (WT1LCs). Within the heart, WT1LCs are predominantly EPDCs6,7, although Wt1CreERT2 also labels other mesothelial cells within the chest, such as the cells of the pericardium and the chest cavity. We pulse labeled adult mice by treating them with tamoxifen, performed LAD ligation to induce MI, and examined hearts 28 days later. Sham operated mice did not develop obvious EAT (0/10), while 53% (10/19) of MI mice were positive for EAT (Fig. 1A,B). Consistent with this observation, myocardial tissue expressed higher levels of adipocyte markers Fabp4, Adiponectin, Adipsin, and Pparg22 (Fig. 1C). The observed adipose tissue fulfilled three criteria for being EAT (Online Supplement, Fig. S1)21: (1) it was perfused from the coronary arteries and not the systemic arteries; (2) it was histologically within the epicardium; and (3) it expressed Ucp1, a marker of brown fat, which was not expressed in peri-aortic fat. In Sham operated mice, immunostaining of tissue sections showed that adipocytes, marked by Perilipin A19,20, were infrequent and rarely overlapped with Wt1LCs. In contrast, after MI adipocytes were more abundant and a subset were Wt1LCs, suggesting that MI stress induces EPDC differentiation into EAT (Fig. 1D).

Fig. 1
MI induces EPDCs to differentiate into epicardial fat

Quantitative analysis showed that 24.1% of Perilipin A + cardiac adipocytes were Wt1LCs (Fig. 1E), consistent with reports that some cardiac adipocytes arise from epicardium1921. The partial labeling of adipocytes in part may reflect incomplete labeling of the Wt1 lineage by tamoxifen, but more likely reflects heterogeous sources for EAT after MI. On the other hand, only a small fraction of Wt1LCs differentiated into adipocytes (Perilipin A +; CD24; 12%) or pre-adipocytes (Perilipin A; CD24+; 3%)19 (Fig. 1F). As a control experiment, we also traced the myocardial lineage using cardiomyocyte specific Myh6-MerCreMer (Myh6-MCM)10. Although EAT was induced by MI with equivalent frequency in Myh6-MCM; Rosa26Tomato mice, we did not observe co-expression of the genetic lineage label in adipocytes (Fig. 1B,D,E). These data indicate that Wt1LCs, but not cardiomyocytes, differentiate into adipocytes in the context of MI.

Epicardial cells in the fetal heart and injured adult heart acquire plasticity by undergoing EMT to form EPDCs.7 Epicardial EMT requires Wt1.23,24 Therefore we hypothesized that EPDC differentiation into adipocytes in the injured adult heart requires epicardial Wt1 expression. Knockout of a conditional Wt1flox allele9 by induction of Wt1CreERT2 in adults caused death from uncertain mechanisms within a few days,25 precluding the use of this model to investigate the functional requirement of Wt1 in EAT formation. To circumvent this problem, we applied Cre modRNA gel8 to the surface of Wt1flox/flox; Rosa26Tomato hearts to locally inactivate Wt1 and simultaneously label EPDCs. In the heart, the Cre gel selectively epicardium and its derivatives8, although it could also label other cell populations that contact the Cre gel including cells lining the pericardium, lungs, and chest cavity. We collectively refer to these as Cre gel lineage cells (CGLCs). In Wt1flox/flox; R26Tomato mice treated with Cre gel, we continued to observe MI-induced formation of EAT (Fig. 1B), consistent with the heterogeneous origin of epicardial fat19,20. However, cells marked by Cre gel did not differentiate into adipocytes, indicating that Wt1 is required for the fate transition of CGLCs into adipocytes (Fig. 1D–F). These results provide functional data that confirm the contribution of Wt1+ mesothelium to EAT.

To further establish the adipogenic potential of EPDCs, we studied the capacity of in vitro cultured primary EPDCs to differentiate into adipocytes. We induced MI in Wt1GFPCre mice, which express GFP-Cre fusion protein from the endogenous Wt1 locus6, and then isolated post-MI EPDCs by fluorescence active cell sorting (FACS) for GFP. When cultured in adipogenic media, the EPDCs differentiated into adipocytes, as demonstrated by Oil red O staining and by upregulation of Fabp4, Adiponectin, Adipsin and PPARγ (Online Supplement, Fig. S2). These results indicate that EPDCs have the potential to differentiate into adipocytes, consistent with our in vivo data and with recent reports19,20.

To investigate the role of paracrine signaling in MI-induced adipogenic differentiation of EPDCs, we first performed microarray expression profiling of post-MI EPDCs cells (marked by GFP in Wt1GFPCre mice) to identify expressed receptors (Fig. 2A). From the microarray data, we selected a panel of 10 canonical candidate receptors expressed in these cells, along with 4 negative controls. To validate the expression of these receptors, we isolated an independent set of post-MI EPDCs. Quantitative reverse transcription PCR (qRTPCR) showed that these cells robustly expressed epicardial markers Wt1 and Tbx1826, and not cardiomyocyte markers Myh6, Tnni3, or Tnnt2 (Fig. 2B), confirming that our isolation procedure captured EPDCs. The 10 candidate receptors identified by microarray were robustly expressed in EPDCs, while the 4 negative control receptors were not (Fig. 2B). We further confirmed expression of these receptors by epicardial cells and their derivatives by immunostaining tissue sections of post-MI Wt1GFPCre hearts (Fig. 2C). Epicardial cells, marked by GFP, co-expressed the 10 candidate receptors. EPDC expression of these receptors was further validated in cultured primary EPDCs (Online Supplement, Fig. S3).

Fig. 2
EPDC paracrine receptor expression

Having identified a number of EPDC-expressed receptors, we next asked if overexpression of their cognate ligands at the time of MI would stimulate EAT formation. To efficiently test these factors for adipogenic activity within the bona fide in vivo context of heart injury, we took advantage of the power of modRNAs to induce myocardial expression of a pulse of paracrine factor.8 We generated a paracrine factor modRNA library encoding the 10 candidate ligands, plus VEGFA, which we previously showed influences EPDC activity8, and luciferase (Luc) negative control. We then scored these factors for adipogenic activity when individually delivered to the heart at the time of MI operation. We found that one ligand, IGF1, robustly stimulated EAT formation after MI so that EAT was observed in 80–100% of hearts, compared to 25–50% at baseline or after treatment with other factors (Fig. 3A). Unlike IGF1, the other 9 paracrine factors did not increase EAT formation after MI above baseline. Interestingly, IGF1 did not stimulate adipogenesis in normal hearts without the stress of MI (Online Supplement, Fig. S4). Increased IGF1-induced adipogenesis was confirmed by measuring expression levels of adipogenic lineage markers Fabp4, Adiponectin, Adipsin, and Pparg (Fig. 3B).

Fig. 3
IGF1 stimulates MI-induced formation of epicardial adipose tissue

Since our data (Fig. 1) and prior studies1921 indicated that EPDCs can differentiate into adipocytes after MI, we cultured post-MI primary EPDCs and treated them with the 10 candidate ligand modRNAs. IGF1, but not the other 9 paracrine factors, stimulated adipogenesis (Fig. 3C). Moreover, qRTPCR comparing cultured EPDCs with or without IGF1 confirmed IGF1 stimulation of EPDC adipogenic differentiation (Fig. 3D). To ask if IGF1 similarly stimulated adipogenesis of human EPDCs, we obtained primary human EPDCs from patients undergoing operation for heart disease. These cells expressed the marker WT1, as expected for EPDCs. Consistent with the data from murine EPDCs, human EPDCs robustly expressed IGF1R (Fig. 3E). Culture of these cells in IGF1 induced adipogenesis to the same extent as adipogenic media, as determined by Oil red O staining (Fig. 3F–G). These data indicate the IGF1-stimulated EPDC adipogenic pathway is conserved to humans.

We next used in vivo lineage tracing strategies to directly measure the effect of each candidate paracrine factor on EPDC adipogenesis in vivo. Fig. 4A shows representative confocal images of the lineage tracing data obtained by Wt1CreERT2-mediated pulse labeling. With control Luc modRNA, we again observed that after MI a small fraction of Perilipin A+ adipocytes co-expressed the genetic marker of Wt1LCs (Fig. 4A, top panel). In contrast, IGF1 modRNA strongly increased the frequency of Perilipin A+ cells in Wt1LCs (Fig. 4A, bottom panel). Quantitative analysis of showed that IGF1, but not the other 9 paracrine factors, increased the fraction of adipocytes that originate from Wt1LCs by about two-fold (Fig. 4B). This was independently supported by Cre gel lineage tracing, which showed a similar increase of adipocytes in the Cre-labeled lineage, although the overall extent of contribution to adipocytes was lower, likely because the gel labels a lower fraction of epicardial cells. EPDCs are a shared and predominant subset of both Wt1LCs and CGLCs. Thus, in combination with our in vitro data, these data suggest that IGF1 augmented the fraction of adipocytes derived EPDCs. While EAT originates from multiple sources, the increase in EPDC-derived adipocytes implies that IGF1 preferentially augments the contribution of EPDCs compared to other sources.

Fig. 4
IGF1 modRNA stimulated adipogenic differentiation of epicardial progenitors

We also analyzed the fraction of EPDCs that express adipocyte or pre-adipocyte markers. IGF1, but not the other paracrine factors tested, strongly increased the fraction of Wt1LCs that differentiate into pre-adipocytes and adipocytes (Fig. 4D), and this was independently supported by the Cre modRNA gel labeling approach (Fig. 4E). The enhanced adipogenic differentiation of Cre-labeled progenitors was accompanied by their significantly reduced differentiation into smooth muscle and endothelial lineages (Online Supplement, Fig. S5). Together these data suggest that IGF1 regulates EPDC fate decisions after MI, directing them towards the adipocyte lineage.

Our finding that IGF1 stimulates adipogenic differentiation suggested that the IGF1 receptor signaling axis is required for MI-induced formation of EAT from EPDCs. To test this hypothesis, we designed at strategy to inhibit IGF1 receptor signaling in EPDCs using either a dominant negative IGF1 receptor mutant27 (IGF1R-DN) or a dominant negative IRS1 mutant28 (IRS1-DN). To validate that these dominant negative proteins block IGF1R phosphorylation, we transfected cultured EPDCs cells with IGF1R-DN, IRS1-DN, or Luc (negative control) and stimulated the cells with IGF1 protein. Western blotting of cell extracts 24 hours later showed that both dominant negative proteins inhibited IGF1R phosphorylation compared to Luc control (Fig. 5A). Furthermore, IGF1-stimulated adipogenic differentiation of cultured EPDCs, measured by oil red O staining, was inhibited 4-fold by the dominant negative mutants (Fig. 5B and Online Supplement. Fig. S6A). This inhibitor effect was further confirmed by qRTPCR measurement of adipogenic lineage marker expression in the EPDC cultures (Online Supplement, Fig. S6B).

Fig. 5
IGF1R signaling is required for EPDC differentiation into adipocytes

To study the requirement of IGF1 receptor signaling in EPDCs for their differentiation towards adipocytes in response to IGF1 modRNA and MI, we delivered the dominant negative IGF1R antagonists to epicardial cells using modRNA gel. Using Wt1CreERT2; Rosa26Tomato lineage tracing mice, we followed the fate of transduced Wt1LCs after MI and IGF1 modRNA treatment. IGF1R-DN or IRS1-DN modRNA gel treatment reduced the frequency of EAT after MI by ~50% (Fig. 5C), reduced the fraction of adipocytes that arise from Wt1LCs (Fig. 5D), and decreased Wt1LC differentiation to pre-adipocytes and adipocytes (Fig. 5E). We confirmed these results using modRNA gel to both Cre label cells derived from epicardium and to concurrently express IGF1R-DN or IRS1-DN. As in the Wt1CreERT2 lineage tracing model, inhibition of IGF1R signaling by either dominant negative mutant in the Cre gel lineage tracing model also reduced adipogenesis of Cre-marked cells in the context of IGF1 modRNA plus MI (Fig. 5F–H). Indeed, the effect of the dominant negative IGF1R inhibitors was slightly more pronounced in the Cre gel lineage tracing model, most likely because most cells that take up one modRNA when co-transfected with multiple modRNAs will concurrently take up all of the modRNAs (Online Supplement Fig. S7). This effect would make lineage-traced and dominant negative inhibitor expressing cells coincide more precisely in the modRNA co-transfection experiment (Fig. 5F–H) compared to separate Wt1CreERT2 labeling and modRNA inhibitor gel expression (Fig. 5C–E). Together both lineage tracing studies indicate that in the post-MI heart exogenous IGF1 signals act on EPDCs via IGF1R to stimulate their adipogenic differentiation.

To determine if the IGF1 signaling axis is required for MI-induced adipogenic differentiation of EPDCs in the absence of exogenously added IGF1 stimulation, we first asked if MI activates the IGF1-IGF1R signaling axis in EPDCs. Immunostaining of peri-infarct myocardium one day after MI showed that IGF1 is upregulated in epicardial cells (Fig. 6A–B). This was further validated by qRTPCR of IGF1 transcripts in peri-infarct tissue, which confirmed marked IGF1 upregulation (Fig. 6C). Immunostaining for activated IGF1R (IGF1R phosphorylated on tyrosine 1161) likewise demonstrated its strong upregulation in EPDCs after MI (Fig. 6D). We investigated hypoxia as a potential trigger for IGF1R activation in EPDCs by culturing EPDCs in normoxic or hypoxic conditions, in the presence or absence of exogenous IGF1 (Online Supplement Fig. S8). In the absence of exogenous IGF1, hypoxia had little effect on total or activated IGF1R. However, addition of exogenous IGF1 in combination with hypoxia increased IGF1R activation, even in the face of decreased total IGF1R. These data indicate that hypoxia strengthens IGF1R activation in EPDCs, suggesting a potential mechanism for the observed interaction between IGF1 and MI in inducing adipogenesis.

Fig. 6
Endogenous IGF1R activation is required for MI-induced EPDC to fat differentiation

Next, we induced MI without addition of IGF1 modRNA, and used modRNA gel to both Cre label EPDCs and to express IGF1R-DN or IRS1-DN, thereby selectively inhibiting IGF1R signaling in Cre gel lineage cells, including EPDCs. IGF1R inhibition reduced the fraction of hearts with grossly evident EAT (Fig. 6E) and reduced expression of adipogenic markers (Fig. 6F). Evaluation of the fate of CGLCs by confocal imaging showed that infrequent Perilipin A+ adipocytes expressed the lineage tracer in post-MI hearts with intact IGF1R signaling (Luc modRNA gel treatment; arrows, Fig. 6G, top panel). Inhibition of IGF1R signaling in CGLCs using either IGF1R-DN or IRS1-DN abolished their differentiation into EAT (Fig. 6G, bottom two panels). This finding was confirmed by quantitative analysis of imaging data, which showed that the fraction of adipocytes derived from CGLCs (Fig. 6H) and the fraction of labeled CGLCs that expressed adipocyte markers (Fig. 6I) were strongly decreased by IGF1R inhibition. Since EPDCs are the predominant constituent of CGLCs, together these data suggest that IGF1R signaling is activated in EPDCs after MI and is required for MI-induced EPDC differentiation into adipocytes.


Our work illustrates the power of modRNA paracrine factor libraries to discover novel in vivo functions of these factors as regulators of progenitor cell fate. By enabling the delivery of pulses of a panel of paracrine factor at the relevant time and place in an in vivo disease context, this approach allows rapid identification of signaling pathways that regulate tissue injury responses in vivo. Here, we used a modRNA paracrine factor screen to identify IGF1 as a factor required for EAT formation.

EAT has an intimate anatomical relationship to the coronary vasculature and myocardium. Epidemiological studies demonstrate that EAT is closely related to coronary artery disease, suggesting that paracrine or metabolic signaling between EAT and the coronary vessels and myocardium influence the evolution of atherosclerotic lesions and heart disease13. Our work shows that IGF1 receptor signaling drives EAT formation by directing progenitor cells into the adipocyte lineage in the context of MI. This function is consistent with the proadipogenic activity of IGF1 on preadipocytes and mesenchymal stem cells.29

The IGF1 pathway has been linked to cardioprotection30 and to enhanced cardiomyocyte differentiation of cardiac progenitor cells.31 Our results delineate potential negative consequences of IGF1 pathway activation, and will need to be considered when evaluating therapeutic interventions based on manipulating the IGF1 pathway. On the other hand, since EAT shares common developmental origins with other types of visceral fat,19 IGF1 may also control expansion of other visceral fat depots. Targeting such signaling pathways or stabilizing mesothelial cells in their epithelial state may be novel approaches to treating the burgeoning problems of ischemic heart disease and obesity.

Consistent with prior studies1921, our work indicated that epicardial progenitors are a source of EAT. This result was supported by two independent lineage tracing strategies, both of which pointed to EPDCs as the IGF1-responsive adipogenic progenitor (Fig. 7): IGF1 receptor activation increased their MI-induced differentiation into adipocytes, while IGF1 receptor inhibition prevented this lineage conversion. The adipogenic effect of IGF1 was conserved in humans EPDCs. Thus, IGF1 signaling is a physiological switch that promotes adipogenic differentiation of epicardial progenitor cells.

Figure 7
Paracrine factor switches guide the fate of cardiac progenitor cells

Importantly, the IGF1-driven progenitor fate switch appeared to be operative only within a brief time window following MI, as IGF1 activation did not stimulate adipogenic differentiation of progenitor cells in the absence of MI, and transient inhibition of IGF1 receptor signaling using modRNA-mediated expression of dominant negative proteins was sufficient to suppress MI-induced differentiation into adipocytes. The requirement for MI in combination with IGF1 to stimulate progenitor differentiation into adipocytes may result from two factors. First, our data suggest that IGF1 more strongly activates IGF1R under hypoxic conditions. Second, MI activation of progenitors may be a prerequisite for their differentiation into adipocytes. MI activates epicardial EMT, which transforms epicardial cells from epithelial cells into mesenchymal cells with greater developmental plasticity7. Supporting this conclusion, Wt1 is required for epicardial EMT,23,24 and MI-induced adipogenesis likewise required Wt1 (Fig. 7).

Infiltration of myocardium by adipocytes is also a hallmark of arrhythmogenic cardiomyopathy, most often caused by mutations in desmosome proteins.38 Presently the origin of this adipose tissue is controversial, and our studies suggest the possibility that desmosomal mutations alter the behavior of epicardial progenitors and direct them to migrate into the myocardium and form adipocytes. The epicardium to endocardium direction of adipogenic involvement in this disease would be consistent with this possibility. Lombardi et al. showed that adipogenic cells in a mouse model of this disease are marked by Isl1Cre39. Although this was interpreted as demonstrating a second heart field origin of the adiogenic cells, Isl1Cre also marks epicardial cells40 and thus the results are consistent with epicardial contribution to adipogenesis in arrhythmogenic cardiomyopathy.

Delineating lineage relationships between populations using Cre labeling approaches is vulnerable to misinterpretation. In this study, we showed that the IGF1R-responsive adipogenic progenitor population is labeled by both Wt1CreERT2 and by Cre modRNA. Within the heart, EPDCs are the population that would be labeled by both techniques. However, the MI model distorts normal tissue planes, removes portions of the pericardium, and can cause cardiac adhesions to surrounding tissues. In this setting, non-cardiac Wt1+ cells from surrounding tissues, such as the mesothelial lining of the chest wall, could contact and infiltrate the heart. Such cells may also be labeled by Cre modRNA gel. The adipocytes in our study had characteristics of EAT, including expression of the brown fat marker UCP1, perfusion by coronary vessels, and covering by epicardium, which excludes mistaking simple pericardial adhesions for EAT. However, we cannot fully exclude the possibility that Wt1+ cells lining the surface of other parts of the chest cavity could have infiltrated EAT and differentiated into EAT-like adipocytes.

Clinical Impact

What is new?

  • We show that a novel paracrine factor library screening approach enables in vivo dissection of signaling pathways that govern cardiac progenitor cell activity.
  • We demonstrate that insulin related growth factor 1 (IGF1) signaling is required in combination with the cellular environment created by myocardial infarction to stimulate differentiation of epicardial cells into adipocytes.

What are the clinical implications?

  • Whereas stimulation of IGF1R may be advantageous for survival of cardiomyocytes and some cardiac progenitors, it may promote formation of epicardial adipose tissue (EAT) in the injured heart.
  • These studies provide the foundation for future mechanistic studies on the pathophyosiolgical significance of EAT.
  • Our results suggest that IGF1R signaling that may be targeted to modulate EAT formation.

Supplementary Material

Online Data Supplement Final


The authors thank V. Huff, University of Texas M.D. Anderson Cancer Center, for the Wt1flox mouse line, and Drs. Marie Jose Goumans and Adriana Gittenberger-de Groot at Leiden University Medical Center for providing human EPDCs.

Funding Sources

W.T.P. was supported by R01 HL094683 and U01 HL100401 from NIH, an Established Investigator Award from the American Heart Association, and by charitable contributions from Dr. and Mrs. Edwin A. Boger. L.Z. was supported by a postdoctoral fellowship from the American Heart Association. K.R.C. was supported by The Wallenberg Foundation, the Karolinska Institute-AstraZeneca Integrated Cardio Metabolic Centre, and a Distinguished Professorship of the Swedish Research Council. We thank Aibin He, Alexander von Gise, and Sean Stevens for their contributions to this project.


Author Contributions

L.Z. designed and performed experiments, analyzed the data, and wrote the manuscript. M.O and L.L.Y. contributed equally. M.O., L.L.Y., and N.S. designed, performed, and analyzed qRT-PCR and immunostaining experiments. L.L.Y also performed animal husbandry and cardiac perfusion. Q.M carried out all mouse surgery. D.S. and Q-D.W. performed experiments on human EPDCs. B.Z performed microarray analysis of EPDCs after MI. W.L.C. performed molecular biology experiments, and A.M performed and analyzed immunostaining data. W.T.P. designed experiments, analyzed data, and wrote the manuscript. K.R.C. contributed to the design of experiments, data review, and manuscript writing.

Competing Interests

K.R.C. is Chair of the External Science Panel for AstraZeneca and Co-Founder of Moderna Therapeutics, which have financial interest in modified RNAs. Daniela Später and Qing-Dong Wang are employees of AstraZeneca.


1. Corradi D, Maestri R, Callegari S, Pastori P, Goldoni M, Luong TV, Bordi C. The ventricular epicardial fat is related to the myocardial mass in normal, ischemic and hypertrophic hearts. Cardiovasc Pathol United States. 2004;13:313–6. [PubMed]
2. Cherian S, Lopaschuk GD, Carvalho E. Cellular cross-talk between epicardial adipose tissue and myocardium in relation to the pathogenesis of cardiovascular disease. Am J Physiol Endocrinol Metab United States. 2012;303:E937–49. [PubMed]
3. Greif M, Becker A, von Ziegler F, Lebherz C, Lehrke M, Broedl UC, Tittus J, Parhofer K, Becker C, Reiser M, Knez A, Leber AW. Pericardial adipose tissue determined by dual source CT is a risk factor for coronary atherosclerosis. Arterioscler Thromb Vasc Biol United States. 2009;29:781–6. [PubMed]
4. Wilm B, Ipenberg A, Hastie ND, Burch JB, Bader DM. The serosal mesothelium is a major source of smooth muscle cells of the gut vasculature. Development. 2005;132:5317–28. [PubMed]
5. von Gise A, Pu WT. Endocardial and epicardial epithelial to mesenchymal transitions in heart development and disease. Circ Res. 2012;110:1628–45. [PMC free article] [PubMed]
6. Zhou B, Ma Q, Rajagopal S, Wu SM, Domian I, Rivera-Feliciano J, Jiang D, von Gise A, Ikeda S, Chien KR, Pu WT. Epicardial progenitors contribute to the cardiomyocyte lineage in the developing heart. Nature. 2008;454:109–13. [PMC free article] [PubMed]
7. Zhou B, Honor LB, He H, Ma Q, Oh JH, Butterfield C, Lin RZ, Melero-Martin JM, Dolmatova E, Duffy HS, Gise A, Zhou P, Hu YW, Wang G, Zhang B, Wang L, Hall JL, Moses MA, McGowan FX, Pu WT. Adult mouse epicardium modulates myocardial injury by secreting paracrine factors. J Clin Invest. 2011;121:1894–904. [PMC free article] [PubMed]
8. Zangi L, Lui KO, von Gise A, Ma Q, Ebina W, Ptaszek LM, Später D, Xu H, Tabebordbar M, Gorbatov R, Sena B, Nahrendorf M, Briscoe DM, Li RA, Wagers AJ, Rossi DJ, Pu WT, Chien KR. Modified mRNA directs the fate of heart progenitor cells and induces vascular regeneration after myocardial infarction. Nat Biotechnol. 2013;31:898–907. [PMC free article] [PubMed]
9. Gao F, Maiti S, Alam N, Zhang Z, Deng JM, Behringer RR, Lecureuil C, Guillou F, Huff V. The Wilms tumor gene, Wt1, is required for Sox9 expression and maintenance of tubular architecture in the developing testis. Proc Natl Acad Sci U S A. 2006;103:11987–92. [PubMed]
10. Sohal DS, Nghiem M, Crackower MA, Witt SA, Kimball TR, Tymitz KM, Penninger JM, Molkentin JD. Temporally regulated and tissue-specific gene manipulations in the adult and embryonic heart using a tamoxifen-inducible Cre protein. Circ Res. 2001;89:20–5. [PubMed]
11. Madisen L, Zwingman TA, Sunkin SM, Oh SW, Zariwala HA, Gu H, Ng LL, Palmiter RD, Hawrylycz MJ, Jones AR, Lein ES, Zeng H. A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat Neurosci United States. 2010;13:133–40. [PMC free article] [PubMed]
12. Muzumdar MD, Tasic B, Miyamichi K, Li L, Luo L. A global double-fluorescent Cre reporter mouse. Genesis. 2007;45:593–605. [PubMed]
13. Tarnavski O, McMullen JR, Schinke M, Nie Q, Kong S, Izumo S. Mouse cardiac surgery: comprehensive techniques for the generation of mouse models of human diseases and their application for genomic studies. Physiol Genomics. 2004;16:349–60. [PubMed]
14. Warren L, Manos PD, Ahfeldt T, Loh YH, Li H, Lau F, Ebina W, Mandal PK, Smith ZD, Meissner A, Daley GQ, Brack AS, Collins JJ, Cowan C, Schlaeger TM, Rossi DJ. Highly efficient reprogramming to pluripotency and directed differentiation of human cells with synthetic modified mRNA. Cell Stem Cell United States. 2010;7:618–30. [PMC free article] [PubMed]
15. Lui KO, Zangi L, Silva EA, Bu L, Sahara M, Li RA, Mooney DJ, Chien KR. Driving vascular endothelial cell fate of human multipotent Isl1+ heart progenitors with VEGF modified mRNA. Cell Res England. 2013;23:1172–86. [PMC free article] [PubMed]
16. da Meirelles LS, Nardi NB. Murine marrow-derived mesenchymal stem cell: isolation, in vitro expansion, and characterization. Br J Haematol England. 2003;123:702–11. [PubMed]
17. Anjos-Afonso F, Siapati EK, Bonnet D. In vivo contribution of murine mesenchymal stem cells into multiple cell-types under minimal damage conditions. J Cell Sci England. 2004;117:5655–64. [PubMed]
18. Bax NA, van Oorschot AA, Maas S, Braun J, van Tuyn J, de Vries AA, Groot AC, Goumans MJ. In vitro epithelial-to-mesenchymal transformation in human adult epicardial cells is regulated by TGFbeta-signaling and WT1. Basic Res Cardiol. 2011;106:829–47. [PMC free article] [PubMed]
19. Chau YY, Bandiera R, Serrels A, Martínez-Estrada OM, Qing W, Lee M, Slight J, Thornburn A, Berry R, McHaffie S, Stimson RH, Walker BR, Chapuli RM, Schedl A, Hastie N. Visceral and subcutaneous fat have different origins and evidence supports a mesothelial source. Nat Cell Biol England. 2014;16:367–75. [PMC free article] [PubMed]
20. Liu Q, Huang X, Oh JH, Lin RZ, Duan S, Yu Y, Yang R, Qiu J, Melero-Martin JM, Pu WT, Zhou B. Epicardium-to-fat transition in injured heart. Cell Res. 2014;24:1367–1369. [PMC free article] [PubMed]
21. Yamaguchi Y, Cavallero S, Patterson M, Shen H, Xu J, Kumar SR, Sucov HM. Adipogenesis and epicardial adipose tissue: a novel fate of the epicardium induced by mesenchymal transformation and PPARγ activation. Proc Natl Acad Sci U S A United States. 2015;112:2070–5. [PubMed]
22. Samulin J, Lien S, Grindflek E, Berget I, Ruyter B, Sundvold H. Depot specific differences during adipogenesis of porcine stromal-vascular cells. Cell Biol Int England. 2008;32:525–31. [PubMed]
23. von Gise A, Zhou B, Honor LB, Ma Q, Petryk A, Pu WT. WT1 regulates epicardial epithelial to mesenchymal transition through beta-catenin and retinoic acid signaling pathways. Dev Biol. 2011;356:421–31. [PMC free article] [PubMed]
24. Moore AW, McInnes L, Kreidberg J, Hastie ND, Schedl A. YAC complementation shows a requirement for Wt1 in the development of epicardium, adrenal gland and throughout nephrogenesis. Development. 1999;126:1845–57. [PubMed]
25. Chau YY, Brownstein D, Mjoseng H, Lee WC, Buza-Vidas N, Nerlov C, Jacobsen SE, Perry P, Berry R, Thornburn A, Sexton D, Morton N, Hohenstein P, Freyer E, Samuel K, van’t Hof R, Hastie N. Acute multiple organ failure in adult mice deleted for the developmental regulator Wt1. PLoS Genet. 2011;7:e1002404. [PMC free article] [PubMed]
26. Cai CL, Martin JC, Sun Y, Cui L, Wang L, Ouyang K, Yang L, Bu L, Liang X, Zhang X, Stallcup WB, Denton CP, McCulloch A, Chen J, Evans SM. A myocardial lineage derives from Tbx18 epicardial cells. Nature. 2008;454:104–8. [PMC free article] [PubMed]
27. Prager D, Li HL, Asa S, Melmed S. Dominant negative inhibition of tumorigenesis in vivo by human insulin-like growth factor I receptor mutant. Proc Natl Acad Sci U S A UNITED STATES. 1994;91:2181–5. [PubMed]
28. Tanaka S, Wands JR. A carboxy-terminal truncated insulin receptor substrate-1 dominant negative protein reverses the human hepatocellular carcinoma malignant phenotype. J Clin Invest UNITED STATES. 1996;98:2100–8. [PMC free article] [PubMed]
29. MacDougald OA, Mandrup S. Adipogenesis: forces that tip the scales. Trends in Endocrinology & Metabolism Elsevier. 2002;13:5–11. [PubMed]
30. McMullen JR, Izumo S. Role of the insulin-like growth factor 1 (IGF1)/phosphoinositide-3-kinase (PI3K) pathway mediating physiological cardiac hypertrophy. Novartis Found Symp England. 2006;274:90–111. [PubMed]
31. Birket MJ, Ribeiro MC, Verkerk AO, Ward D, Leitoguinho AR, den Hartogh SC, Orlova VV, Devalla HD, Schwach V, Bellin M, Passier R, Mummery CL. Expansion and patterning of cardiovascular progenitors derived from human pluripotent stem cells. Nat Biotechnol. 2015;33:970–979. [PubMed]
32. Smart N, Risebro CA, Melville AA, Moses K, Schwartz RJ, Chien KR, Riley PR. Thymosin beta4 induces adult epicardial progenitor mobilization and neovascularization. Nature. 2007;445:177–82. [PubMed]
33. Tian X, Pu WT, Zhou B. Cellular Origin and Developmental Program of Coronary Angiogenesis. Circ Res. 2015;116:515–530. [PubMed]
34. Smart N, Bollini S, Dube KN, Vieira JM, Zhou B, Davidson S, Yellon D, Riegler J, Price AN, Lythgoe MF, Pu WT, Riley PR. De novo cardiomyocytes from within the activated adult heart after injury. Nature. 2011;474:640–4. [PMC free article] [PubMed]
35. Zhou B, Pu WT. Genetic Cre-loxP Assessment of Epicardial Cell Fate Using Wt1-Driven Cre Alleles. Circ Res. 2012;111:e276–80. [PMC free article] [PubMed]
36. Christoffels VM, Grieskamp T, Norden J, Mommersteeg MT, Rudat C, Kispert A. Tbx18 and the fate of epicardial progenitors. Nature. 2009;458:E8–9. discussion E9–10. [PubMed]
37. Rudat C, Kispert A. Wt1 and epicardial fate mapping. Circ Res. 2012;111:165–9. [PubMed]
38. Iyer VR, Chin AJ. Arrhythmogenic right ventricular cardiomyopathy/dysplasia (ARVC/D) Am J Med Genet C Semin Med Genet United States. 2013;163C:185–97. [PubMed]
39. Lombardi R, Dong J, Rodriguez G, Bell A, Leung TK, Schwartz RJ, Willerson JT, Brugada R, Marian AJ. Genetic fate mapping identifies second heart field progenitor cells as a source of adipocytes in arrhythmogenic right ventricular cardiomyopathy. Circ Res. 2009;104:1076–84. [PMC free article] [PubMed]
40. Zhou B, von Gise A, Ma Q, Rivera-Feliciano J, Pu WT. Nkx2-5- and Isl1-expressing cardiac progenitors contribute to proepicardium. Biochem Biophys Res Commun. 2008;375:450–3. [PMC free article] [PubMed]